DESCRIPTION (Derived in part from the applicant's abstract): Haemophilus influenzae continues to be a significant health pathogen, causing airway mucosal infections in developed countries and pneumonia in developing nations. The noncapsular determinants of H. influenzae virulence are poorly understood. The applicant has previously demonstrated that human hemoglobin binds to the surface of a heme-restricted H. influenzae via a 120 kDa protein. This protein was cloned using the oligonucleotides corresponding to the end-terminal amino acid sequence, and a gene hgpA, encoded hemoglobin binding activity in E. coli. The nucleotide sequence of hgpA revealed homology to iron regulated genes in other genera and indicated a putative outer membrane leader sequence, a tonB region, and a series of CCAA repeats. To examine the role of hemoglobin binding in the pathogenesis of H. influenzae disease, the applicant will clone and characterize hgpA from H. influenzae strain 1388. (The gene has been cloned form strain 689.) Through search of the H. influenzae RD genome database, the applicant has found four ORFs with homology (ranging from 49% identity and 65% similarity, to 84% identity and 90% similarity) in that organism. He will clone the homologs from his two pathogenic isolates: strain 689 and strain 1388. The function of the genes in E. coli and phenotypic changes resulting from insertional inactivation will be studied. In Specific Aim #2, he'll characterize the role of the CCAA repeats in the regulation of expression of hgpA and the other homologous hemoglobin binding genes in the two wild-type isolates. He will determine the transcription start side of hbpA and the regulation through frame shift by RT-PCR of middle ear effusions from children with otitis media. In addition the individual clones will be sequenced to determine the number of CCAA repeats. To investigate the role of the hemoglobin proteins in pathogenesis, the applicant will use two methods: 1) a competitive reverse transcriptase polymerase chain reaction which will be applied to both samples obtained from human infections as well as in animal models. In addition, he will investigate the pathogenic potential of the two wild-type strains, and mutants in each of the genes encoding hemoglobin binding proteins for virulence in the infant rat model of invasive disease, and the chinchilla model of otitis media.